We have been interested in building a mathematically-robust framework to detect magic/nonstabilizerness in quantum devices, in the overall interest of understanding how to probe quantum complexity.

In Witnessing Magic with Bell inequalities, we propose to use Bell inequalities to detect magic in multipartite quantum systems, finding them to witness a strong form of magic, referred as "non-local magic".  On Prepare-and-Magic,  we also discussed how magic can be exploited to have higher success probabilities of quantum random access codes, leading to different witnesses on Prepare-and-Measure scenarios.

Another line of my research investigates how fermionic states map to qubit states under local post-selection—a process of measuring local fermionic charges and applying feedback operations. This framework provides a new lens for understanding fermionic systems under very strong local interactions. In Partons from stabilizer codes, we made this connection precise for fermionic stabilizer states, demonstrating that after projection, they are universally described by topological phases in the same class as the toric code. Intriguingly, our results also provide sharp counter-examples to long-standing expectations in the condensed matter theory of strongly interacting electrons.